UKH

The coefficient of friction of ice/ice is 0.01. I assume the ice is dry.

The coefficient of friction for most metals ranges from about 0.30 to 0.60 so ice is 30 to 60 times slidier.

Is it something to do with hydrogen bonds?

Heard an explanation that ice crystals cannot bond properly on the top layer of the ice sheet. So the result will be similar to a sheet of ice with loose molecules on top which will make it very slippery.

Not sure if this is right or scientifically proven but sounds reasonable to me.

I always thought it was dependent on temperature and relied on the contact pressure being enough to melt the surface so you slide on water trapped between the ice and whatever is sliding on it - go cold enough and it isn't very slippery?

The thing is it doesn't seem to matter how cold it is if you step on a frozen puddle without looking you always end up on your arse.

If they don't bond to each other very well on the surface why do they bond very well inside the bulk of the ice?

Is there any solid that's slippier than ice? If there is then it must be pretty esoteric because I don't think I've come across anything.

On the surface there aren't enough 'contact points' to form a strong bond between the crystals. Inside the bulk all the crystals are surrounded on all sides by other crystals which will be sufficient points of contact to form a very solid structure.

The top layer of the ice will be a continuous process of loose ice crystals bonding and breaking free so to say.

Disclaimer: This could all be scientifically incorrect. I'm not a scientist.

There isn’t really one “coefficient of friction” for a material.  The coefficient depends upon both of the two materials in contact and upon each of their surface roughnesses.  Surface roughness in itself is a complicated multi-dimensional parameter.  That’s before you start considering the ability of surfaces to modify each other, eg polishing or melting.

Naturally formed ice tends to be very smooth, and it’s tendency to melt and re-freeze keeps it that way.  Snow is “just” ice but it has a much tougher surface and is less slippery.

Even if the ice is dry, at the microscopic contact points the ice melts due to the high pressure. Hence no ice to ice contact, just ice to water to ice contact.

Only the high points on the surfaces are in contact. The rest of the surface is dry and solid.

That is a suggestion. I don't know if it is true.

Edit - a quick look online suggests it is more complicated than that

People used to think the low friction was caused by a thin layer of water that melted at the surface due to the applied pressure, but I think that explanation is out of favour now.  Applying pressure does lower the melting temperature of ice, but the numbers don't really work.

What people now believe is that it is due to the phenomenon of "surface melting" - there is always a microscopically thin layer of water at the surface, even below the melting temperature of the ice.  

(Incidentally, this is why you can supercool water well below 0 C, but you can't superheat ice above 0 C.  To freeze water you always need a nucleation site, but to melt ice the surface water layer just grows into the crystal).

This was always my understanding. The top layer is active relative to temperature, extremely cold and the molecules are less active, so it's less slippy. 

The surface melting under contact can't be right, as the pressure applied would need to extreme, elephant in stilettos scenario. 

Interesting idea.

Rambling diversion...

I once left a lager too long in the freezer when I wanted a cold beer quickly.

Nothing happened when opened but when I swigged from the bottle it turned into slush in a top to bottom wave. Mucky mouth I reckon.

An ice surface - solid, liquid or gas? Below surface - solid, above surface - gas at SVP, at surface - ??? 

 I suppose if water with bonds to neighbours in all directions is solid at 0C then surface molecule with neighbours and bonds in half as many directions isn't so solid.

Ice Abhors a Vibraam

Doesn't completely answer your question, but as you're asking, I think you'll find it interesting...

youtube.com/watch?v=wMFPe-DwULM&

The bit about ice starts around 1 minute in, but frankly the whole video is worth a watch, with one of the greatest physicists of the 20th century. 

I've always understood cartilage to be more slippery than ice and a quick google seems to confirm that but with answers ranging between 3 and 1000 times more slippery!

Perhaps it's how you measure it, or which type of cartilage you're comparing with which type of ice?

Some other contenders mentioned in this discussion

https://arstechnica.com/civis/viewtopic.php?t=574520

Yes, doesn't take much to set off supercooled water, so you may not even have been that mucky.

Considering the necessarily broken bonds at the surface of an ice crystal is a good way of thinking about it - that means that a bare ice crystal surface has rather a high energy.  If you put a thin layer of liquid water there that smooths the transition between the ordered molecules in the crystal and the air/vapour.  To be more formal, if the surface energy of the bare crystal is bigger than the sum of the ice/water interfacial energy and the surface energy of water then the system will lower its energy by sticking in a thin layer of water between the ice and the vapour, even below 0 K.

My understanding is that cartilage (in vivo, anyway) is slippery because it's coated with a "hairy" layer of long molecules, which trap a layer of water.  So when two bits of cartilage come into microscopic contact they're kept a few 10's of nm apart by these water-swollen, hairy layers, meaning they can run past each other very smoothly.

Try grabbing hold of a piece of metal that has been well frozen and covered with a thin layer of ice - not so slippy then!

..and here's a link to a detailed explanation: https://physicstoday.scitation.org/doi/10.1063/1.2169444

see fig 5 for example.

although I'm still not quite sure what happens below about -33 deg C where the layer of water disappears. Given that no one has ever reported that ice loses its slipperiness at extremely low but naturally occuring temperatures I'm thinking that the water molecules on the surface must still be poorly bonded t the bulk of the ice...but I think someone has already said that.

The obvious explanation for the disappearance of the layer below -33 would be that at that temperature the energy penalty of having liquid water is finally too large for even the gain in the surface energies to counteract.  But nothing is simple with water and I wonder whether the fact that -30 is about the temperature at which supercooled liquid water turns into a glass might be relevant too.

Edit: just remembered, that last bit is wrong, the glass transition temperature for pure water is quite a bit lower.

You can cool liquids below their nominal freezing point, but they will not solidify efficiently unless there is a seed that allows crystals to form. Opening the bubble lowered the pressure, which created bubbles, which served as seeds for the ice crystals.

CB

No, it was my mucky mouth as nothing happened until I took a sip.

I'll need to do some repeat and control experiments now.

I get that but as far as I know there isn't a step change in coefficient of friction around that temperature...and if there isn't is the water really the major reason for the low coefficient of friction?

I don't think the water layer disappears abruptly at -33, doesn't its thickness go to zero continuously, so rather than expecting an abrupt transition in friction at -30, wouldn't you'd just expect it to increase continuously as the surface water layer became thinner and so less and less effective as a lubricating layer?

I don't know.  I do know that these water layers are a real pig to measure, though.

At some point it must disappear I assume. Or at least be so thin that it can no longer act as a lubricant?

I thought I'd have a look for a graph of coefficient of friction vs. temperature and I've found one here but it doesn't give a clear answer: https://www.tribonet.org/wiki/friction-coefficients-in-atmosphere-and-vacuu...

While there is a kink in the graph of ice on ice at about -27 deg C the coefficient of friction actually *decreases*. The curve for ice on ice is linear all the way to about -90 deg C.

Curious.

Yes, I think you would expect the layer to disappear at low enough temperature.  But I suspect that no-one has actually measured it very well.  Because you can't do the measurements in vacuum (because the presence of the layer is crucially dependent on there being an equilibrium with water vapour) you have to use optical techniques, which rely on the (very small) difference in refractive index between water and ice, and are going to be struggling for layers whose thickness is a very small fraction of the wavelength of light.

Friction is nothing if not complicated!  Ultimately it arises from energy dissipation around the contact points, there may be a number of different mechanisms for this whose relative importance varies with temperature in quite a complex way.

Is that not how ice skates work though?

You also need to factor in the effect of friction & pressure on the ice melting it and forming a thin layer of "lubricating" water

Not from my understanding. Even footwear that spreads the load still slips, unless you have studs of some type that grip through the loose molecular layer, crampons, studded winter tyres etc. 

Skis are wide, you could have a big pair of fat powder skis and they'll slide until you tip them and let the edge bite in. The same way you stop on skates, or in fact push forwards. 

Be careful leaving cans of lager in the freezer; when it freezes hard the ice increases the internal volume and creates huge pressure so when you open the can the unfrozen ethanol liqueur with all the colour and flavour will burst out and soak your ceiling with a sticky brown mess. And it tastes even worse than diluted lager. Guess how I know this?

a guy from alaska says: When it gets to minus 40 degrees Fahrenheit (minus 40 degrees Celsius), the snow "becomes like sandpaper." Truffer's observation lines up with what the Bonns found. At ultralow temperatures, the molecules on the surface don't have as much energy to break and create bonds as they roll around, so the ice becomes nonslippery

https://www.livescience.com/62621-why-is-ice-slippery.html

[Bonn describes the layer as a two-dimensional gas rather than a three-dimensional liquid..]

i'm surprised the https://en.wikipedia.org/wiki/Van_der_Waals_force hasn't been mentioned. surely it has a role to play?

Yes but the description refers to snow, not ice and contradicts the graph I posted of friction coefficient vs. temperature which shows a fairly continuous increase as temperature decreases.

In The Worst Journey in the World, Apsley Cherry Garrard describes pulling sledges across Antarctica in the middle of winter and likens it to pulling them through sand. My take on this is that it is so cold and dry the the snow flakes aren't bonding together and aren't melting. Years ago I designed a set of antennas to go on a telecoms line right up the West coast of Greenland. The most challenging region that had the worst environment was the Southern tip where icing and very high winds were the norm. In the North it was so cold that the snow behaved like dust. It didn't stick to anything and when coming to rest in a cavity or corner, didn't turn into a block if ice.

maybe he was referring to compacted snow. i do like the idea of ice like sandpaper

You can't measure it in a vacuum, because the change in pressure would change the melting point of the water. 

There probably wouldnt be any ice, just water vapour, unless you were at less than -70 odd degrees.

I seem to remember our mad monk Physics teacher calling that phenomenon regelation(?) back in the early 70s